I was a young scientist back when "Industrial Revolution 2.0" was about nanotechnology in the early 00's... You know, actual atoms, not bits. Since then that term has been used mostly in the digital space, where it is (IMHO) vastly misapplied.

I have no doubt that AI will provide an important advancement of society in many ways. But I'm sorry, LLM's aren't going to cure cancer or do anything truly "unimaginable"... Precisely because their training set is the "average" set of knowledge of humanity. They're trained precisely *not* to think out of the box, because they have no logical faculties.

Modern (LLM) AI's inadequacy to do anything "industrial" is laughably demonstrated by the following: Ask your favorite AI image generator to make a hexagonal grid. Nothing fancy, no cats in spacesuits, no slop: Just a simple grid of dots in a hexagonal pattern.

It will fail drastically, every time. Because it's a stochastic generator of things, not a deterministic one.... And *this* is what's going to build the factories?